A physics-informed neural network for creep-fatigue life prediction of components at elevated temperatures

Xiao-Cheng Zhang,Jian-Guo Gong,Fu-Zhen Xuan

DOI: https://doi.org/10.1016/j.engfracmech.2021.108130

IF: 5.4

2021-12-01

Engineering Fracture Mechanics

Abstract:Physics-informed neural network has strong generalization ability for small dataset, due to the inclusion of underlying physical knowledge. Two strategies are enforced to incorporate physics constraints to a deep neural network in this work. One is to obtain extended features through physics-informed feature engineering, and the other is to incorporate physics-informed loss function into deep neural network as constraints. Conventional machine learning models, deep neural network and physics-informed neural network are applied to predict creep-fatigue life of 316 stainless steel. Results show that physics-informed neural network presents better prediction accuracy than deep neural network and conventional machine learning models.

mechanics